sign in sign up
<<
Home , Cigarette, Cigarette pack, Longbeach (cigarette), Morley (cigarette), Plain tobacco packaging

Research Tob Control: first published as 10.1136/tc.2008.026732 on 30 September 2008. Downloaded from How does increasingly plainer packaging influence adult smokers’ perceptions about image? An experimental study M A Wakefield, D Germain, S J Durkin

Centre for Behavioural Research ABSTRACT For example, British American (BAT) in , The Cancer Council Background: Cigarette packaging is a key marketing experimented with its design Victoria, Victoria, Australia strategy for promoting brand image. Plain packaging has on packs of Benson and Hedges and 4 Correspondence to: been proposed to limit brand image, but tobacco in 2002–3 and introduced split Dr M A Wakefield, Centre for companies would resist removal of branding design packs (so-called ‘‘kiddie packs’’) in 2006,56 by Behavioural Research in Cancer, elements. which two low-consumption smokers could The , 1 Rathdowne Street, Carlton, Method: A 3 (brand types) 6 4 (degree of plain easily procure and split apart a single pack for their Victoria, Australia 3053; packaging) between-subject experimental design was own use. Some have also begun to melanie.wakefield@cancervic. used, using an internet online method, to expose 813 incorporate the colour schemes of graphic health org.au adult Australian smokers to one randomly selected warnings into the overall colour and design of the , after which respondents completed entire pack, causing the warnings to become less Received 9 July 2008 Accepted 28 September 2008 ratings of the pack. salient since they in with the overall pack Published Online First Results: Compared with current cigarette packs with full design (Kylie Lindorff, Quit Victoria, personal 30 September 2008 branding, cigarette packs that displayed progressively communication, July 2008). Bans on traditional fewer branding design elements were perceived increas- forms of tobacco advertising and promotion also ingly unfavourably in terms of smokers’ appraisals of the lead to a more critical role for cigarette packaging packs, the smokers who might smoke such packs, and at the point of sale, where packs are designed to the inferred experience of a cigarette from these allow brand families to better stand out at the cash packs. For example, cardboard brown packs with the register.27 These point-of-sale tobacco advertising number of enclosed cigarettes displayed on the front of and cigarette displays create an enticing in-store the pack and featuring only the brand name in small presence for youth,8–10 and a cue to prompt adult standard font at the bottom of the pack face were rated smokers to purchase.11 as significantly less attractive and popular than original In response to these developments, proposals to branded packs. Smokers of these plain packs were rated introduce ‘‘plain’’ cigarette packaging have http://tobaccocontrol.bmj.com/ as significantly less trendy/stylish, less sociable/outgoing emerged whereby packs would be stripped of and less mature than smokers of the original pack. colours, brand imagery, corporate logos and trade- Compared with original packs, smokers inferred that marks and manufacturers would be permitted to cigarettes from these plain packs would be less rich in print only the brand name in a mandated size, font tobacco, less satisfying and of lower quality tobacco. and location, in addition to required health Conclusion: Plain packaging policies that remove most warnings and other legally mandated information brand design elements are likely to be most successful in such as toxic constituents, tax seals or pack removing cigarette brand image associations. contents.12 13 Aside from denying that the pack is a of advertising, a key argument of the against plain packaging is that it on September 24, 2021 by guest. Protected copyright. In the face of comprehensive restrictions on would amount to trademark infringement and tobacco advertising and promotion, tobacco packa- unjustifiably encumber the use of in ging has become the primary vehicle for commu- the course of trade, violating several international nicating brand image.1 Through the use of colour, trade and intellectual property agreements such as fonts, images and trademarks, cigarette packs the Trade-Related Aspects of International project a brand image that says something about Property Rights (TRIPS) Agreement 1994, the the user of the product. Commonly referred to as a North American Trade Agreement 1994 ‘‘badge product’’, the user often associates with the (NAFTA) and the Paris Convention for the identity and personality of the brand image.23 Protection of Industrial Property 1883.13–15 Unlike most other consumer products, cigarette However, as Freeman and colleagues13 argue, the packs remain with users once opened and are industry’s interpretation of these agreements is repeatedly displayed in social situations, thereby selective, as each of these treaties contains specific serving as a direct form of mobile advertising for exemptions allowing necessary measures to be the brand. adopted to protect public health and to protect the In countries such as Australia where traditional public interest. forms of advertising are banned, packaging now Research by the tobacco industry has shown This paper is freely available serves as the vehicle for tobacco marketing. that the design of a cigarette pack can not only online under the BMJ Journals unlocked scheme, see http:// Accordingly, Australian tobacco companies have generate powerful images about the type of person tobaccocontrol.bmj.com/info/ experimented with producing more colourful and who might typically smoke the brand, but also unlocked.dtl varied packs, as well as designs to pique curiosity. provide cues about the sensory perceptions of the

416 2008;17:416–421. doi:10.1136/tc.2008.026732 Tob Control: first published as 10.1136/tc.2008.026732 on 30 September 2008. Downloaded from Research paper smoke which may be expected from a particular cigarette. For Procedure example, given identical cigarettes to try, men and women rated Eligible participants in the panel were sent an email that the sensory experience of smoking a cigarette differently included a web link to the survey, inviting them to participate depending on the brand name given to the cigarette, with in a study about their opinions of a brand with which they women rating the attributes of the smoke more positively when might be familiar. Respondents were given a chance to win one assigned a feminine brand name and men rating it more of 10 AU$100 shopping vouchers as an incentive to participate. positively if it had a masculine brand name.16 Similarly, sensory A reminder email was sent 5 days after the initial email, and a perceptions of cigarettes can be manipulated simply by final reminder was sent a further 5 days later. Upon accessing changing the colour or shade of colour on a pack, through a the survey website, demographic information was collected process called ‘‘sensation transfer’’. for including sex, age, level of educational attainment, postcode and Filter cigarettes revealed that increasing the amount of white whether they were daily or weekly smokers of manufactured space on the pack and lightening brown colour tones reduced cigarettes. Respondents who said they smoked less than weekly the perception of the cigarette’s strength when the cigarette was or not at all and/or those outside the age criteria were excluded smoked.17 Research conducted by Philip Morris USA also from further participation in the study. indicated strong sensation transfer effects when testing Eligible respondents were then randomly allocated to view identical Ultra Light cigarettes placed in either a blue one of 12 pack conditions that varied by brand and extent of or red pack. Although the cigarettes were exactly the same, plain packaging. The three brands were the three most popular those placed in the red packs were perceived to be ‘‘harsher’’ Australian brand variants among adult smokers (Winfield Blue 25 than those in the blue packs, while cigarettes in the blue packs 25s; Peter Jackson Rich 30s; Rich 40s). Previous were rated as ‘‘too mild’’, ‘‘not easy drawing’’ and ‘‘burned too tobacco company research on packaging perceptions has found fast’’.18 that particular pack colours are associated with specific Previous experimental studies examining the potential impact perceptions—for example, red connotes strength in taste, blue suggests a strength cigarette and white connotes the of plain packaging have shown that health warnings are more 23 19–22 freshest and lightest cigarettes of all. As much is already noticeable when presented on a plain cigarette pack, and known about the effects of specific pack colours, the current that plain packs detract noticeably from brand imagery study did not test different pack colours but presented all plain established by cigarette brands.20 21 23 To our knowledge, no packs in a cardboard brown colour previously demonstrated to research has examined the effects of plain packaging on elicit negative responses.26 27 The four pack design conditions smoker’s perceptions of taste, strength or quality of the product, were: and little attention has so far been focused on the testing of c Original pack: an existing pack one could purchase today. different plain pack versions against each other, examining the c Plain pack 1: a generic cardboard brown pack that maintains impact of branded fonts and other brand elements on packs. a branded font (ie, original font size, style and position) and This study aims to provide research evidence to assist the positioning of brand/descriptor. selection of plain pack designs that would promote the least http://tobaccocontrol.bmj.com/ c Plain pack 2: a generic cardboard brown pack with the brand positive attributes about smoking for smokers. We hypothesise name in a standard font in a prominent position on the pack that smokers will rate an original branded pack more positively with descriptor information in a standard font at the than their plain pack counterparts, and that plain packs with bottom. progressively fewer brand-associated elements will be rated c Plain pack 3: a generic cardboard brown pack with the brand more negatively. name in a smaller standard font positioned at the bottom and ‘‘(xx number) cigarettes’’ in a larger font in a prominent METHODS position on the pack. All pack conditions had the same graphic health warning Design visible on the top of the face of the pack as required by This study employed a 3 (brand types) 6 4 (degree of plain Australian Government legislation.28 In light of the tobacco packaging) between-subject experimental design using an on September 24, 2021 by guest. Protected copyright. industry’s argument that enforcement of plain packaging would internet online method to expose adult smokers to one amount to trademark infringement and unjustifiably encumber randomly selected cigarette pack, after which respondents the use of trademarks in the course of trade, during the completed ratings of the pack. development of our hypotheses and the designs of generic packs for testing, legal advice from an intellectual property lawyer was Sample sought to ensure that we would be testing packs that could A market research company was commissioned to undertake realistically be introduced into the market place without the administration of the survey. A sampling frame of adults impeding trademark laws. Figure 1 displays each of the 12 pack aged 18–49 years was sourced from an existing national online conditions. panel. The panel members were originally sourced from various After viewing their assigned pack, respondents completed methods including computer-assisted telephone interviews and ratings of the pack in relation to perceived attributes of the face-to-face market research, during which participants supplied brand, perceived attributes of smokers of the brand and their email address and gave permission to be contacted by expected taste/quality of the cigarette. The assigned pack was email to participate in future research as well as through online present on the screen as the smoker completed each of the marketing and other online databases. The panel was broadly ratings. representative of Australian Bureau of Statistics norms in relation to geographical location, income and age. Using Questionnaire Cohen’s power calculations,24 we estimated that a sample size Attributes to be rated were modified from past tobacco industry of 780 would allow the detection of small-to-medium effect packaging studies where smokers were asked to rate cigarette sizes for main effects (,0.50; p = 0.05; power = 0.99). packs on attractiveness, brand imagery characteristics and

Tobacco Control 2008;17:416–421. doi:10.1136/tc.2008.026732 417 Research paper Tob Control: first published as 10.1136/tc.2008.026732 on 30 September 2008. Downloaded from

Figure 1 Original and plain packs for each brand.

perceived sensory attributes.29 30 In the current study, respon- participation and told they had been entered in the draw for the http://tobaccocontrol.bmj.com/ dents were asked to rate the cigarette pack they were shown in shopping vouchers. relation to: brand image (the mental associations that are stimulated by the pack’s appearance alone); smoker attributions Statistical analysis (anticipated personality/character type of the typical person Analysis of variance and x2 tests were used to check that who might be expected to regularly smoke the pack displayed); random assignment yielded equivalent groups with respect to and inferred smoking experience (the type of smoking experi- smoking history and demographic characteristics. Preliminary ence which might be anticipated from a cigarette contained in analyses indicated that survey responses on the 11-point the displayed pack). response scale were not normally distributed. Responses were When viewing the cigarette pack, respondents were asked to skewed at two points on the scale: at 0 (indicating disagree- rate the following phrases describing attributes of the cigarette ment) and at 5 (indicating moderate agreement). We therefore pack shown from 0 (not at all well) to 10 (extremely well). dichotomised responses to permit statistical analysis, with on September 24, 2021 by guest. Protected copyright. ‘‘This pack …’’: ‘‘is a popular brand among smokers’’; ‘‘has an responses from 0 to 4 categorised together to reflecting attractive looking pack’’; ‘‘is good value for money’’; ‘‘is an ‘‘disagreement to low agreement’’ and responses from 5 to 10 exclusive/expensive brand’’; and ‘‘is a brand you might try/ reflecting ‘‘moderate to high agreement’’. Differences between smoke’’. Looking at the same pack, respondents were then asked pack conditions were assessed using logistic regression analysis to rate a number of attributes of typical smokers of the pictured to generate odds ratios and confidence intervals. cigarette pack from 0 (not at all) to 10 (extremely well). ‘‘A typical smoker of this pack is …’’: ‘‘trendy/stylish’’; ‘‘young’’; ‘‘masculine’’; ‘‘lower class’’; ‘‘sociable/outgoing’’; ‘‘older/ RESULTS mature’’; and ‘‘confident/successful’’. Finally, looking at the Sample characteristics and group assignment same pack, respondents were asked to think about how a Overall, 813 regular smokers resident in Australia completed the cigarette from the pictured pack might taste, and to rate the study procedure, yielding a response rate of 22% of all those sent following descriptions on how well they relate to the pack email invitations. In total, 62% of smokers were female, 81% shown from 0 (not at all) to 10 (extremely). ‘‘These cigarettes were aged 30 years or older, 36% had completed Year 11 would taste …’’: ‘‘rich in tobacco flavour’’; ‘‘low in and secondary education or less, 45% had completed Year 12 ’’; ‘‘of cheap tobacco’’; ‘‘satisfying’’; ‘‘like a light education or some tertiary, and 19% had completed a tertiary cigarette’’; ‘‘of the highest quality tobacco’’; and ‘‘harsh on qualification. Just under half (47%) smoked .15 cigarettes per the throat’’. Within each of the questions, attributes were day on average. Respondents were also classified by postcode of presented randomly to avoid order effects. residence into four levels of social advantage/disadvantage based Once the final question was completed, respondents sub- on the Socio-Economic Index for Areas (SEIFA) developed by mitted their responses to the survey, were thanked for their Bureau of Statistics.31 Just under one-quarter

418 Tobacco Control 2008;17:416–421. doi:10.1136/tc.2008.026732 Tob Control: first published as 10.1136/tc.2008.026732 on 30 September 2008. Downloaded from Research paper

Table 1 Demographic and smoking characteristics of participants by pack condition Original Plain pack 1 Plain pack 2 Plain pack 3 (n = 176) (n = 219) (n = 199) (n = 219) p Value

Male (%) 38.6 38.4 35.2 40.2 0.765 Age (%) 0.206 18–29 years 18.2 17.4 24.6 17.8 30+ years 81.8 82.6 75.4 82.2 Education (%) 0.684 Year 11 or less 31.7 36.8 35.6 39.0 Year 12/some tertiary 50.6 42.0 45.2 43.0 Tertiary 17.7 21.2 19.1 18.0 Socioeconomic status (%) 0.409 SEIFA 1 (lowest advantage) 23.3 19.6 20.6 20.8 SEIFA 2 18.2 21.0 18.1 20.4 SEIFA 3 30.7 36.5 37.2 27.3 SEIFA 4 (highest advantage) 27.8 22.8 24.1 31.5 Consumption (%) 0.355 1–10 cigs/day 29.0 27.9 24.6 27.9 11–15 cigs/day 24.4 29.7 22.1 26.9 16–20 cigs/day 25.0 16.9 21.1 22.8 21–25 cigs/day 11.4 13.7 14.6 10.5 26 + cigs/day 10.2 11.9 17.6 11.9 Brand seen is brand smoked 19.3 17.4 15.6 15.5 0.727 SEIFA, Socio-Economic Index for Areas.

(21%) of respondents lived in areas of low advantage, while 27% Effect of pack condition on perceptions were living in areas of high advantage. Overall, 17% of The results of fitting a logistic regression model with an participants were assigned to view a brand that they smoked. interaction between pack condition and brand to predict pack Table 1 shows that demographic and smoking characteristics of perceptions indicated that there were no interactions between the respondents did not vary significantly across the different these two variables. Therefore, in the following analyses, the pack conditions. An average of 203 respondents (minimum 176; results for the three brands were aggregated. Table 2 shows that maximum 219) were randomly allocated to each of the four for all brands combined, Plain pack 1, which preserved the

pack conditions. placement and font of brand names and brand variants, was http://tobaccocontrol.bmj.com/

Table 2 Bivariate logistic regression analyses comparing percentage of smokers who agreed with rated attributes, by pack condition{ Original Plain pack 1 Plain pack 2 Plain pack 3 % OR % OR % OR % OR OR for linear trend

Brand/pack characteristics Popular brand among smokers 83.5 1 78.1 0.70 75.9 0.62{ 67.1 0.40*** 0.75*** Attractive looking pack 50.0 1 34.7 0.53** 31.2 0.45*** 32.0 0.47*** 0.79*** Value for money 56.8 1 55.7 0.96 50.8 0.78 49.3 0.74 0.90{ Exclusive/expensive brand 39.8 1 44.7 1.23 38.2 0.94 40.2 1.02 0.97

Brand you might try/smoke 59.1 1 55.7 0.87 53.3 0.79 51.6 0.74 0.91 on September 24, 2021 by guest. Protected copyright.

Smoker characteristics Trendy/stylish 47.2 1 38.4 0.70{ 34.2 0.58* 32.0 0.53** 0.81** Young 55.1 1 52.1 0.88 41.2 0.57** 47.9 0.75 0.88* Masculine 58.0 1 59.8 1.08 55.8 0.92 42.9 0.55** 0.81*** Lower class 52.8 1 54.3 1.06 50.3 0.90 53.0 1.01 0.99 Sociable/outgoing 68.8 1 55.7 0.57** 51.8 0.49*** 49.3 0.44*** 0.78*** Older/mature 67.0 1 65.8 0.94 61.8 0.80 55.7 0.62* 0.85* Confident/successful 51.7 1 51.6 1.00 42.7 0.70 43.4 0.72 0.87*

Perceived sensory perceptions Rich in tobacco 76.1 1 70.8 0.76 64.8 0.58* 67.1 0.64* 0.86* Low in tar and nicotine 44.9 1 38.4 0.76 33.7 0.62* 33.3 0.61* 0.85* Tastes of cheap tobacco 54.5 1 47.0 0.74 50.3 0.84 50.7 0.86 0.97 Satisfying 72.7 1 65.3 0.71 64.8 0.69 61.2 0.59* 0.86* Like a light cigarette 47.2 1 41.1 0.78 43.2 0.85 39.7 0.74 0.92 Of the highest quality tobacco 60.8 1 59.8 0.96 51.8 0.69{ 50.7 0.66* 0.85* Harsh on throat 50.6 1 48.9 0.93 54.3 1.16 52.5 1.08 1.05 *p,0.05; **p,0.01; ***p,0.001. {Scored 5 or more on a scale from 0 (not at all) to 10 (extremely). {p,0.10.

Tobacco Control 2008;17:416–421. doi:10.1136/tc.2008.026732 419 Research paper Tob Control: first published as 10.1136/tc.2008.026732 on 30 September 2008. Downloaded from

brown colour are perceived increasingly unfavourably by smokers. Even though all plain packs substituted a cardboard brown colour for the original pack colour, the removal of additional design elements produced measurable decrements in smokers’ appraisals of the packs, the smokers who might smoke such packs, and the inferred experience of smoking a cigarette from these packs. Although we did not explicitly test this, it is possible that the gradual removal of design elements may also have served to increase the salience of the pictorial health warnings as suggested in earlier research,19–22 and this would be a desirable additional outcome. There are a number of study limitations that should be mentioned. First, the use of an 11-point response scale produced an irregular response distribution and we needed to dichotomise responses to conduct analysis. In future studies a more usual 5- point Likert scale with named response options would be preferred. However, even though we dichotomised responses, we were still able to detect differences between pack conditions. Second, although we tested three variations of plain packs, each condition removed several design elements at one time and we Figure 2 Smokers’ ratings by pack condition. were not able to determine which specific brand elements most contributed to deteriorations in smoker perceptions of the perceived as less attractive than the original branded pack, and packs. Other study designs such as fractional factorial design smokers of the pack were perceived as less sociable and outgoing where a single brand element can be manipulated may be better than smokers of the original pack. There was also a trend for suited for this more finely-tuned purpose.32 However, our study smokers of Plain pack 1 to be perceived as less trendy and stylish has shown that, in aggregate, smokers perceive plain cardboard than smokers of the original pack. On all other dimensions, brown packs with fewer branding elements less favourably, and Plain pack 1 was rated as similar to the original branded pack. this applied to the three brand variants most commonly smoked Compared with the original branded pack, Plain pack 2, in Australia. Along the same lines, we may have obtained which standardised the placement and font of the brand name different results using packs with different background colours and relinquished the brand variant to standard type at the other than the cardboard brown we selected. However, the bottom of the pack, was rated as less attractive, and smokers of colour selected was chosen purposively as a result of previous 26 the brand were rated as less trendy and stylish, less young and research where it elicited negative perceptions. Third, our http://tobaccocontrol.bmj.com/ less sociable and outgoing. In addition, compared with those study displayed packs via an internet image which did not who viewed the original pack, fewer smokers who viewed Plain permit smokers to handle the pack. This reduction in pack- pack 2 thought the cigarettes would be low in tar, fewer related information might have been expected, however, to thought the cigarettes would be rich in tobacco and of the understate the brand design elements, leading to underestimates highest quality tobacco. There was also a tendency for Plain of differences between pack conditions. Thus, our study results pack 2 to be rated as less popular than the original pack. may be conservative. In addition, confidence in the validity of Compared with the original branded pack, Plain pack 3, responses would have been stronger if a rationale was provided where the brand name and variant appeared only in small to respondents for the existence of the plain packs. Finally, the standard type at the bottom of the pack, was perceived as being internet method of survey administration may have allowed less popular and less attractive, and smokers of the brand were some smokers to seek the input of others into the responses perceived to be less trendy and stylish, less masculine, less they gave. However, if this occurred, the randomised design on September 24, 2021 by guest. Protected copyright. sociable or outgoing and less mature. Compared with those who would have meant that this kind of interference in responses viewed the original pack, significantly fewer smokers who was equally distributed across conditions. As our sample was viewed Plain pack 3 thought the cigarettes would be low in tar, sourced from an existing online panel with a consequent low rich in tobacco, satisfying to smoke and of the highest quality response rate, respondents were not representative of the tobacco. Table 2 also shows that, for most of these mentioned What this paper adds attributes, there was a significant linear decline in the degree of favourable ratings as pack branding design information reduced. To graphically represent this (fig 2), we combined the variables c has been proposed as a means to within each of the three categories of ratings (ie, brand/pack limit brand imagery, but little research has been undertaken to characteristics; smoker characteristics; perceived sensory percep- guide decision-making about which packaging brand design tions) after testing the strength of correlations within each elements drive brand appeal for smokers. c category (brand/pack characteristics: Cronbach’s a = 0.72; This experimental study found that plain packs with smoker characteristics: Cronbach’s a = 0.87; perceived sensory increasingly fewer brand design elements are perceived perceptions: Cronbach’s a = 0.74). increasingly unfavourably in terms of smokers’ appraisals of the packs, the smokers who might smoke such packs, and the inferred experience of smoking a cigarette from these packs. DISCUSSION c This implies that tobacco control policies should aim to This study suggests that cigarette packs that display progres- remove as many brand design elements as possible. sively fewer branding design elements and presented in a generic

420 Tobacco Control 2008;17:416–421. doi:10.1136/tc.2008.026732 Tob Control: first published as 10.1136/tc.2008.026732 on 30 September 2008. Downloaded from Research paper general population in terms of demographic characteristics. 6. Anon. Cigarette split pack defeated. Daily Telegraph 18 Nov 2006, Sydney, Australia. 7. Pollay R. More than meets the eye: on the importance of retail cigarette However, this was an experimental study rather than a merchandising. Tob Control 2007;16:270–4. population survey, and the online method was simply used to 8. Henriksen L, Flora JA, Feighery E, et al. Effects on youth of exposure to retail recruit smokers to the experiment and randomise them to one tobacco advertising. J Appl Soc Psychol 2002;32:1771–89. 9. Henriksen L, Feighery EC, Wang Y, et al. Association of retail tobacco marketing of the experimental conditions. Randomisation was successful with adolescent smoking. Am J Public Health 2004;94:2081–3. as judged by the fact that groups did not differ in composition. 10. Wakefield M, Germain D, Durkin S, et al. An experimental study of effects on Overall, our internet method of stimulus presentation provided schoolchildren of exposure to point-of-sale cigarette advertising and pack displays. a simple inexpensive experimental method for obtaining Health Educ Res 2006;21:338–47. 11. Wakefield M, Germain D, Henriksen L. The effect of retail cigarette pack displays on responses from a large sample size to randomly-presented impulse purchase. Addiction 2008;103:322–8. stimulus packs. 12. Cunningham R, Kyle K. The for plain packaging. Tob Control 1995;4:50–86. With a likely acceleration in the rate of comprehensive 13. Freeman B, Chapman S, Rimmer M. The case for plain packaging of tobacco products. Addiction 2008;103:580–90. restrictions on tobacco advertising and promotion as countries 14. Philip Morris. Draft insert re: generic packaging. USA: Philip Morris, 1999. Bates No. strive to meet their responsibilities under the Framework 2064831501/1505. http://legacy.library.ucsf.edu (accessed 7 Jul 2008). Convention on Tobacco Control (FCTC),33 tobacco packaging 15. Cunningham R. Plain packaging. In: Smoke and mirrors: the Canadian tobacco wars. will assume even greater importance internationally as a International Development Research Centre, Ottawa , 1996. 3 16. Freedman HH, Dipple WS. The effect of masculine and feminine brand names on the promotional vehicle for driving brand image. Plain packaging perceived taste of a cigarette. Decision Sci 1978;9:467–71. measures remain an important yet relatively under-explored 17. Etzel E. Consumer research proposal: Camel Filter revised packaging test study.RJ component of tobacco control legislation designed to compre- Reynolds. 2 Mar 1979. Bates No. 500566627/6632. http://legacy.library.ucsf.edu (accessed 7 Jul 2008). hensively eliminate all forms of tobacco advertising and 18. Isaacs J. Identified HTI test of Marlboro Ultra Lights in a blue pack vs. Marlboro Ultra 13 promotion. In their review, Freeman and colleagues conclude Lights in a red pack. USA: Philip Morris, July 1981. Bates No. 2047387079/7089. that trademark laws and international trade laws do not http://legacy.library.ucsf.edu (accessed 7 Jul 2008). preclude mandating the removal of brand design elements on 19. Beede P, Lawson R. The effect of plain packages on the perception of cigarette health warnings. Public Health 1992;106:315–22. tobacco packs and that plain packaging could and should be 20. Goldberg ME, Liefeld J, Kindra G, et al. When packages can’t speak: possible pursued under the FCTC. Our research extends the existing impacts of plain and generic packaging of tobacco products: expert panel report. evidence base by demonstrating not only that plain packs are Prepared for Health Canada, 1995. 21. Rootman I, Flay B. A study on plain packaging, health warnings, event perceived unfavourably by smokers, but that plain packs with marketing, and price reductions key findings. Toronto: University of Toronto, Centre for the least brand design elements have the least appeal. Further Health Promotion, 1995. research to quantify more carefully the effects of specific design 22. Goldberg ME. The effect of plain packaging on response to health warnings. elements on brand perceptions—including among youth at risk Am J Public Health 1999;89:1434. 23. Centre for Health Promotion. Effects of plain packaging on the image of tobacco for smoking—would provide helpful guidance for future policy products among youth. Prepared for the Canadian Cancer Society, 30 Nov 1993. development. 24. Cohen, J. Statistical power analysis for the behavioural sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, 1988. Acknowledgements: The authors thank Matthew Rimmer and Jonathan Liberman 25. Nielsen AC. Top 100 brands. Retail World 2007;60:88–90. for advice on legal aspects of tobacco packaging. 26. Centre for Behavioural Research in Cancer. Adolescents’ reactions to cigarette http://tobaccocontrol.bmj.com/ packs modified to increase extent and impact of health warnings. In: Health warnings Funding: This study was funded by Quit Victoria and the Cancer Council Victoria. and contents labelling on tobacco products. Prepared for the Australian Ministerial MAW was supported by an Australian National Health and Medical Research Council Council on Drug Strategy Tobacco Task Force on Tobacco Health Warnings and Principal Research Fellowship. Content Labelling. Melbourne: Anti-Cancer Council of Victoria, 1992. 27. Trachtenberg JA. Here’s one tough cowboy. Forbes Magazine 9 Feb 1987:109–11. Competing interests: None. 28. Department of Health and Ageing (DHA). Health warnings on tobacco product Ethics approval: The study was approved by the Human Research Ethics Committee packaging: trade practices (Consumer Product Information Standards) of the Cancer Council Victoria. (Tobacco) Regulations 2004, Dec 2007. http://www.health.gov.au/internet/main/ publishing.nsf/Content/health-pubhlth-strateg-drugs-tobacco-warnings.htm (accessed 8 Jul 2008). REFERENCES 29. Philip Morris. Consumer tracking study: Philippines, 28 December 1988. Tobacco 1. National Cancer Institute. The role of the media in promoting and reducing tobacco Industry Gateway Documents. Bates: 2504007423/2504007477. use. Tobacco Control Monograph No 19. Bethesda, MD: US Department of Health 30. Philip Morris. Cigarettes questionnaire, September 1989. Tobacco Industry Gateway and Human Services, National Institutes of Health, National Cancer Institute, 2008: Documents. Bates No. 2504034620/4625. http://legacy.library.ucsf.edu (accessed 7 on September 24, 2021 by guest. Protected copyright. 106–8. NIH Publication No 07-6242. http://cancercontrol.cancer.gov/tcrb/ Jul 2008). monographs/19/index.html. 31. Australian Bureau of Statistics (ABS). Information paper: 2001 census of 2. Wakefield M, C, Horan JK, et al. The cigarette pack as image: new evidence population and housing: socio-economic indexes for areas. Catalogue No. 2039.0. from tobacco industry documents. Tob Control 2002;11(Suppl I):i73–80. Canberra: ABS, 2003. 3. Hammond D. FCTC Article 11: Tobacco packaging and labelling: a review of 32. Hammond D. The impact of brand descriptors, package design and emission evidence. University of Waterloo, Canada, Nov 2007. http://www.cctc.ca/cctc/EN/ information on risk perception. Presented at Symposium on Advances in Tobacco tcrc//tcmonograph.2007-12-19.7863543963 (accessed 7 Jul 2008). Product Labelling and Health Warnings: Evidence to Inform FCTC Article 11. 14th 4. Wakefield M, Letcher T. My pack is cuter than your pack. Tob Control Annual Meeting of the Society for Research on Nicotine and Tobacco, Portland, 2002;11:154–6. Oregon, 27 Feb–1 Mar 2008. 5. Chapman S. Australia: ‘addresses’ youth smoking. Tob 33. World Health Organization. WHO Framework Convention on Tobacco Control. Control 2007;16:2–3. 2005. http://www.who.int/fctc/en (accessed 7 Jul 2008).

Tobacco Control 2008;17:416–421. doi:10.1136/tc.2008.026732 421